Meta paid contractors to pose as children, flood rival chatbots with prompts about suicide, eating disorders, and abuse, then log every response in spreadsheets. The company calls it "industry-standard safety benchmarking," but the operation had no consent, no disclosure, and no shared findings: the four things that make red teaming legitimate. Host Emily Laird walks through the Wired investigation, the 45,000-prompt testing rounds, and the court testimony showing what Meta knew about its own failure rates while it was busy documenting everyone else's. This is the difference between a shield and a sword, and the paperwork says sword.────────────

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

📩 EMAIL REMINDERS FOR THE MEETUPS

https://app.e2ma.net/app2/audience/signup/2101263/1779703/

💬 CONNECT WITH EMILY LAIRD ON LINKEDIN

http://www.linkedin.com/in/meet-emily-laird

Host Emily Laird breaks down why chatbots so often agree with your worst instincts, then shows how a pre-mortem prompt turns that people-pleasing machinery against your plan. The stakes are practical: job offers, house purchases, program launches, hard conversations, and every other moment when agreement feels comforting but costs you later. This episode is a reality check on AI sycophancy, decision stress-testing, and the simple question that can make a yes-machine finally tell you what might fail.──────────────

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

📩 EMAIL REMINDERS FOR THE MEETUPS

https://app.e2ma.net/app2/audience/signup/2101263/1779703/

💬 CONNECT WITH EMILY LAIRD ON LINKEDIN

Microsoft cut 4,800 jobs, and the savings cover roughly two days of its AI infrastructure spending. Host Emily Laird runs the arithmetic the press release skipped: a $190 billion capex bill, an Xbox division losing 64 cents on every dollar, and a $625 billion backlog where nearly half the money traces back to one cash-burning customer. This is the story of a company growing 18 percent while shedding a trillion dollars in market value. The layoffs were never a savings plan, they were a message.

──────────────────

📑READ DEAN BALL'S ESSAY

https://www.hyperdimensional.co/p/what-should-be-done

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

📩 EMAIL REMINDERS FOR THE MEETUPS

You now know more about Microsoft's recent layoffs than you did before you arrived.

OpenAI just offered the US government a 5% ownership stake, and host Emily Laird checks the receipts on what that $42.6 billion gift actually costs. This episode traces the timeline from frozen model releases to a confidential IPO filing, and explains why handing equity to your regulator looks less like patriotism and more like the most expensive insurance policy in corporate history. From the Alaska Permanent Fund plumbing to Bernie Sanders wanting ten times more, the whole deal gets priced out in plain English. It's a reality check on who owns what when the referee asks to join the team.

──────────────────

📑READ DEAN BALL'S ESSAY

https://www.hyperdimensional.co/p/what-should-be-done

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

📩 EMAIL REMINDERS FOR THE MEETUPS

You now know more about OpenAI's repoted government gift than you did before you arrived.

Elon Musk says Grok 4.5 rivals Claude Opus, but there's no public benchmark, no model card, and no outside lab backing the claim. Host Emily Laird peels the branding off and looks at what actually matters: a closed-loop AI machine built from SpaceX engineers, Tesla codebases, Cursor data, and reinforcement learning that never stops running. The real contest isn't model versus model, it's feedback loop versus feedback loop. Find out why the loudest claim in the announcement might be the least important part.

──────────────────

📑READ DEAN BALL'S ESSAY

https://www.hyperdimensional.co/p/what-should-be-done

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

📩 EMAIL REMINDERS FOR THE MEETUPS

You now know more about Grok 4.5 than you did before you arrived.

In this episode, host Emily Laird exposes the massive delusion behind corporate AI bans and the quiet rise of off-the-books model usage. Managers believe a strict firewall stops unauthorized tech, but it only forces employee innovation into the gray market of Shadow AI. Instead of pretending the technology does not exist, you will get a precise blueprint for building a one-page rulebook your team will actually respect. It is time to replace reactive boardroom panic with transparent and auditable parameters.

──────────────────

📑READ DEAN BALL'S ESSAY

https://www.hyperdimensional.co/p/what-should-be-done

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

📩 EMAIL REMINDERS FOR THE MEETUPS

You now know more about managing Shadow AI than you did before you arrived.

In this episode, host Emily Laird breaks down Dean Ball’s argument that frontier AI safety is shifting from public frameworks to quiet release control. The real stakes are not just whether advanced models are risky, but who gets to decide when they are safe enough to use. This is a sharp look at government pressure, lab accountability, and why regulating the model may miss the system that actually matters.

──────────────────

📑READ DEAN BALL'S ESSAY

https://www.hyperdimensional.co/p/what-should-be-done

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

📩 EMAIL REMINDERS FOR THE MEETUPS

You now know more about Dean Ball's Essay than you did before you arrived.

OpenAI’s GPT-5.6 Sol is not just another smarter model, it is a preview of controlled intelligence as the new product category. In this episode, host Emily Laird breaks down the real stakes: government-visible model releases, cyber capability, activation classifiers, trusted-access programs, and the uncomfortable fact that frontier models are now being watched while they run. The hype says “better AI,” but the actual story is stranger, colder, and more consequential: access, safety, cost, and control are becoming part of the product itself.

──────────────────

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

📩 EMAIL REMINDERS FOR THE MEETUPS

You now know more about GPT 5.6 Sol than you did before you arrived.

In this episode of Generative AI 101, host Emily Laird breaks down recursive self-improvement, the feedback loop where AI begins helping build better AI. The reality check: this is not a chatbot fantasy, it is frontier labs using models to write code, run tests, build evaluations, and speed up the next generation of systems. When Anthropic says Claude authored more than 80 percent of merged code in its own codebase, the question stops being theoretical and starts becoming operational. This episode is about what happens when AI becomes part of the production line for AI itself.

──────────────────

🎯 JOIN THE AI WEEKLY MEETUPS

https://www.uwstout.edu/ai-weekly-meetup

📩 EMAIL REMINDERS FOR THE MEETUPS

You now know more about recursive self-improvement than you did before you arrived.

AI scams have gone full Death Star, scaling panic with cloned voices, deepfakes, fake romances, and corporate video-call trickery. Host Emily Laird breaks down how fraud became a factory in 2026, from virtual kidnappings to pig butchering bots with the charm of a Tinder date and the soul of a parking ticket. The fix is gloriously low-tech: slow down, call back, use code words, and never let urgency do your thinking for you.